Situation models and children’s reading comprehension: what role does visual imagery play?

Individual differences in children’s reading comprehension have been attributed to the level at which a reader is able to construct a coherent meaning-based mental representation of the situation described in a text (i.e., a “situation model”). However, although there is evidence that situation models contain perceptual information such as visual imagery, it is yet to be established whether visual imagery contributes to children’s reading comprehension via its role in situation model construction. To investigate this, three studies were conducted with children in Grades 4 and 5 (age range: 8.08-11.17 years) as part of the current thesis. Study 1 explored the utility of several measures of visual imagery and examined whether this construct is best captured by the differentiation of separate visual imagery processes in this younger population. Fifty-nine children completed five measures of visual imagery, each designed to capture a distinct subcomponent of the visual imagery system, including image generation, image maintenance, image scanning, image transformation, and image strength/vividness. It was found that the visual imagery measures were not highly related to one another and thus each represented a unique construct. However, not all of the included measures proved to be valid and reliable. Utilising the measures of visual imagery that were found to have adequate psychometric properties in Study 1, Study 2 then examined the influence of different subtypes of visual imagery (image maintenance, image scanning and image transformation) on individual differences in reading comprehension. In addition, this study further investigated existing criticisms that traditional measures of reading comprehension do not capture all of the skills involved in situation model construction, by including two separate measures of reading comprehension: a traditional standardised measure (the Neale Analysis of Reading Ability), and a newer measure designed from cognitive theory, which measures higher-level comprehension processes separate to the effects of lower-level reading ability (the Diagnostic Assessment of Reading Comprehension; DARC). It was found that each subtype of visual imagery differentially predicted reading comprehension. In addition, each measure of reading comprehension was differentially influenced by variations in word reading ability and verbal working memory, with evidence that the Neale was more influenced by lower-level reading skills and simple verbal working memory, whereas the DARC was more influenced by non-verbal reasoning and complex verbal working memory. However, visual imagery was not found to be a reliable predictor of reading comprehension; although, this may have been due to an incongruity between the type of imagery that occurs during objective tasks of visual imagery and the visual simulation of narrative events. Thus, Study 3 was designed to disrupt good and poor comprehenders’ visual imagery during reading in order to determine whether good comprehenders show more reliance on visual imagery during comprehension than poor comprehenders. Unexpectedly, however, good comprehenders showed limited evidence of engaging in higher-level comprehension processes (i.e., predictive inferencing) even when imagery was not impaired. Despite this, important implications regarding the use of both textbase and imagery-based representations were revealed, as poor comprehenders displayed increased difficulty maintaining a verbal load during reading compared to a visuospatial load. This suggests that in comparison to good comprehenders, poor comprehenders may have a greater reliance on textbase over imagery-based representations during reading. Overall, this thesis adds to the literature that suggests not all reading comprehension measures are interchangeable in regard s to the underlying skills that they measure. Further, visual imagery may be relevant to reading comprehension; yet, it is likely that this relationship will be further established through careful conceptualisation and measurement of visual imagery versus visual simulation. These findings have implications regarding the use of existing comprehension measures in research and practice, and may also aid future research that investigates the role of visual imagery in higher-level comprehension processes